Ascorbic acid, corticosteroids, and thiamine in sepsis: a review of the biologic rationale and the present state of clinical evaluation

Thiamine (vitamin B1) is a water-soluble vitamin that is a key component of a number of cellular metabolic processes. In its phosphorylated form, thiamine pyrophosphate, thiamine acts as a cofactor for pyruvate dehydrogenase, the enzyme necessary for converting pyruvate to acetyl-coenzyme A for entry into the Krebs cycle. When thiamine levels are insufficient, pyruvate is unable to be converted to acetyl coenzyme A, resulting in impaired aerobic respiration and a compulsory shift to the anaerobic pathway, resulting in elevated serum lactate levels [26‐28]. Thiamine also plays a role in the metabolism of branched-chain amino acids and is a critical component of the pentose phosphate pathway, which is essential for the generation of NADPH and therefore glutathione cycling, an important anti-oxidant pathway [29‐31]. A thiamine deficiency syndrome, beriberi, bears a number of similarities to sepsis, including peripheral vasodilation, cardiac dysfunction, and elevated lactate levels [29].

Thiamine deficiency is not rare in critically ill populations and may be associated with increased mortality in some cases [7, 29, 32, 33]. Further, thiamine levels are depleted during the course of critical illness and the administration of thiamine during critical illness may improve organ dysfunction [34‐36]. In a mouse model of cardiac arrest, the provision of thiamine improved mitochondrial function, reduced histologic signs of brain injury, and improved neurologic outcomes [37]. In a canine model of septic shock, thiamine pyrophosphate improved lactate clearance, oxygen consumption, and arterial pressure irrespective of thiamine deficiency status [38]. In the only randomized trial of thiamine in human septic shock, 88 patients were selected for increased risk of symptomatic thiamine deficiency based on a serum lactate > 3 mmol/L after volume resuscitation. In that study, there was no effect overall of thiamine on the primary outcome of median lactate level at 24 h, though there was a statistically significant difference when evaluating lactate levels at serial timepoints in the first 72 h. In a pre-defined subgroup of patients with thiamine deficiency (35% of the cohort), however, the administration of thiamine reduced lactate levels and improved mortality [7]. In a post hoc analysis of that study, patients without baseline end-stage renal disease who were given thiamine had better renal outcomes than those randomized to placebo [8]. Thiamine supplementation has not been shown to be associated with significant adverse effects, even at high doses [39]. Overall, these results suggest a role for thiamine supplementation as a low-risk and potentially high reward intervention for some patients with septic shock and increased baseline risk of thiamine deficiency.

Like thiamine, ascorbic acid is a water-soluble vitamin essential for a number of processes in the human body. As an anti-oxidant, ascorbic acid is an electron donor that directly scavenges free radicals, prevents the generation of new free radicals through its suppressive effects on the NADPH oxidase (NOX) pathway, and assists in the recycling of other anti-oxidants [40‐42]. The anti-oxidant effect of ascorbic acid results in reduced endothelial permeability, improved microvascular and macrovascular function, and attenuated cellular apoptosis in pathological states [41, 43, 44]. In addition, ascorbic acid has a number of effects on the immune system, including regulation of macrophage function, reduction of inflammatory mediators, and even a direct bacteriostatic effect at high concentrations [45‐47]. Lastly, ascorbic acid is essential in the generation of endogenous vasopressors and may be important in maintaining vascular vasopressor responsiveness [41, 48].

Prior studies have found that plasma and cellular levels of ascorbic acid decline rapidly during critical illness [49‐51]. Similar to thiamine, ascorbic acid deficiency syndrome (scurvy) bears a number of similarities to sepsis, including malaise, coagulation abnormalities, and endothelial wall breakdown [49]. While interest in ascorbic acid for the management of critical illness has recently been reinvigorated, it is not new. The potential benefit of ascorbic acid for reducing resuscitation fluid requirements in burn patients and organ dysfunction in critically ill surgical patients was suggested over a decade ago [52, 53]. More recently, small randomized trials in sepsis have shown promise. In one study, 24 septic patients were randomized in a 1:1:1 ratio to receive high dose ascorbic acid (200 mg/kg), low dose ascorbic acid (50 mg/kg), or placebo. In that study, no adverse effects were related to the ascorbic acid and patients who received ascorbic acid had more rapid reduction in measures of organ injury, inflammation, and procalcitonin. There also appeared to be a dose–response relationship, with patients who received higher dose ascorbic acid having more rapid clinical improvement [9]. In a more recent study, 28 patients with vasopressor-dependent septic shock were randomized to 25 mg/kg of ascorbic acid every 6 h or placebo [10]. Those in the ascorbic acid arm required lower vasopressor doses and had lower mortality.

While the high doses of ascorbic acid given in the above clinical studies were not associated with any identified harms specific to the drug, one theoretical concern regarding the routine use of ascorbic acid in sepsis is the potential for increased oxalate excretion and the development of oxalate renal calculi [54]. Thiamine pyrophosphate is a key co-enzyme necessary for the function of glyoxylate aminotransferase, which catalyzes the breakdown of glyoxalate to carbon dioxide instead of oxalate. Thiamine deficiency states, therefore, may predispose to increased oxalate excretion [54, 55]. It should be further noted that short-term, intravenous ascorbic acid—even at high doses—has not been found to increase the risk of renal calculi in controlled trials to date [41, 43]. Other potential adverse effects of vitamin C include abdominal pain/bloating, increased iron absorption, hemolysis in patients with G6PD enzyme deficiency, and false negative results on fecal occult blood testing [56]. At very high doses, ascorbic acid may act as a pro-oxidant, although this has not been found to be the primary effect in vivo [41, 57]. Finally, high doses of ascorbic acid may falsely elevate glucose level readings when measured with certain point-of-care glucometers employing glucose dehydrogenase-pyrroloquinoline quinone amperometric methods [58].

A number of large, randomized trials have assessed the added benefit of corticosteroids when included as part of general septic shock management. These studies have generally shown corticosteroids to improve various clinical outcomes in septic shock (e.g., time to shock reversal, ventilator-free days), but there have been mixed results with respect to mortality [59‐62]. Whether the routine administration of hydrocortisone to patients with septic shock should be standard remains a matter of debate [11, 12].

The biologic basis for the inclusion of hydrocortisone in the drug combination, however, is based on potential synergy between ascorbic acid and hydrocortisone. Glucocorticoid binding to glucocorticoid receptors is negatively affected by oxidizing molecules. This may be reversed by the administration of ascorbic acid, which has been shown to restore glucocorticoid receptor function [13]. The cellular uptake of ascorbic acid is mediated by the sodium-vitamin C transporter (SVCT2), which is downregulated during inflammatory states. The administration of glucocorticoids has been shown to increase expression of the transporter [14, 15]. In a study examining the barrier function of human lung microvascular epithelial cells, the combination of ascorbic acid and hydrocortisone showed a synergistic barrier-protective effect after lipopolysaccharide exposure—above the combined effect of either agent when given alone [15].

The combination of thiamine, ascorbic acid, and corticosteroids has been studied in two, single center, before-and-after cohort studies [6, 63]. In addition to the above-referenced study by Marik et al., a recent study performed in South Korea compared 53 patients with severe pneumonia admitted to the ICU who received the thiamine, ascorbic acid, and hydrocortisone combination to historical controls. In that study, patients who received the thiamine, ascorbic acid, and hydrocortisone combination had a substantial mortality benefit (adjusted odds ratio 0.15, 95% CI 0.04–0.56). Although there were baseline imbalances in the ‘control’ and ‘treatment’ groups wherein patients in the ‘treatment’ group were more likely to have been receiving vasopressor and renal replacement therapy, the mortality benefit persisted after propensity-adjustment and propensity-matching. While these studies were the first to explore the drug combination in severe infection, their observational methodology, inclusion of non-consecutive and non-concurrent ‘control’ arms, small sample sizes, and single center nature represent significant limitations and preclude broad conclusions regarding the efficacy of this drug combination in sepsis.